213-240_Beukes

8/20/2019 213-240_Beukes

1/28

COPPER ELECTROWINNING: THEORETICAL AND PRACTICAL DESIGN 213

Introduction

The electrowinning of copper ions derived from leaching, or solvent extraction is a significant

contributor to the global copper commodity supply. The process of electrolysis for copper was

first developed in the late 19th century and despite numerous advancements in technology the

principles and basic equipment remain the same. The first part of this paper deals with the

theoretical requirements and fundamental equations and principles that govern copper

electrowinning. The second part discusses the practical requirements for designing a copper

electrowinning plant.

BEUKES, N.T. and BADENHORST, J. Copper electrowinning: theoretical and practical design.

Hydrometallurgy Conference 2009, The Southern African Institute of Mining and Metallurgy, 2009.

Copper electrowinning: theoretical and practical

design

N.T. BEUKES* and J. BADENHORST*

*TWP Matomo Process Plant, South Africa

An engineering house’s perspective of required inputs in designing a

copper electrowinning tank house and ancillary equipment calls for

both understanding of the key fundamental controlling mechanisms

and the practical requirements to optimize cost, schedule and productquality. For direct or post solvent extraction copper electrowinning

design, key theoretical considerations include current density and

efficiency, electrolyte ion concentrations, cell voltages and electrode

overpotentials, physical cell dimensions, cell flow rates and electrode

face velocities, and electrolyte temperature. Practical considerations

for optimal project goals are location of plant, layout of tank house

and ancillary equipment, elevations, type of cell furniture, required

cathode quality, number and type of cells, material of construction of

cells, structure and interconnecting equipment, production cycles,

anode and cathode material of construction and dimensions, cathode

stripping philosophy, plating aids, acid mist management, piping

layouts, standard electrical equipment sizes, electrolyte filtration,

impurity concentrations, bus bar and rectifier/transformer design,

electrical isolation protection, crane management, sampling and

quality control management, staffing skills and client expectations,.

All of the above are required to produce an engineered product that

can be designed easily, constructed quickly and operated with

flexibility.

8/20/2019 213-240_Beukes

2/28

HYDROMETALLURGY CONFERENCE 2009214

The hardware used is simple in nature; for electrowinning an acid resistant bath withanodes and cathodes submersed in solution with current passing through the electrodes is thefundamental process unit. The fundamental concepts lie in reaction kinetics, mass transferphenomenon, thermodynamics and other electrochemical specific models, the application of which leads us to a deeper and more appreciative knowledge of the ‘simple’ electrowinningreactor.

The first part of the paper goes through the fundamentals and culminates in an examplereactor being developed. Note that not all design procedures are named as this wouldcompromise TMPs intellectual property. However the reader will be able to get a very goodunderstanding of what is required to design and build a copper electrowinning plant.

Part 1: Copper electrolysis theoretical considerations

Faraday’s law

For the winning of copper by the addition of electrons

[1]

Cations go towards the cathode, and anions go to the anode. The working electrode is wherereduction takes place and the counter electrode is where oxidation occurs. The workingelectrode is the cathode and the counter electrode the anode. For the generaloxidation/reduction reaction:

[2]

Faraday’s Law gives the total amount of charge spent to reduce M mols of Ox (Q) is:

[3]

The charge spent per unit time is defined as the current (I):

[4]

Normalizing with unit area gives Faraday’s Law expressed in Current Density (i) :

[5]

Faraday’s law then is: the current flowing in an external circuit is proportional to the rate of the reaction at the electrode.

Nernst equation

The standard electrode potential is the potential difference between energy states of product

and reactant and is a manipulation of the Gibbs free energy reaction (G)Reaction thermodynamics gives the following relationship for Gibbs Free energy:

[6]

For a single electrode

[7]

Since the electrode potential regulates the energy of electron exchange, it also controls thecurrent and thus the rate of exchange. Current and Potential (E) are dependant variables of one

another. Where the work done (W) is related to the Potential difference by:

8/20/2019 213-240_Beukes

3/28


[8]

Substituting for (Q) and work (W) for (G), where (W) and (G) are in joules then Gibbs FreeEnergy can be re-written as:

[9]Substituting (7) and (8) with (9) gives the Nernst Equation for an electrode (or half cell):

[10]

Mechanism of electron transfer

For elementary reactions at an electrode the following two mechanisms are primarilyresponsible for electron transfer.

Mass transfer controlled

1) Diffusion of copper cations from the bulk phase to where the reaction occurs at thesurface.

[11]

Reaction kinetics controlled

2) Heterogeneous transfer of electrons from the solid electrode to the copper cation at thesurface of the electrode.

[12]

Further phenomena, coupled chemical reactions, adsorption and formation of phases arereported to also have a role in the electron transfer mechanism. The formation of phases isrelevant to the plating of copper on the cathode and involves nucleation and crystal growthsteps. Copper atoms diffuse through the solid phase to a location in an appropriate site of thecrystal lattice. Adsorption and nucleation steps are considered to be included in theHeterogeneous Electron Transfer reaction rate mechanism.

The overall rate is controlled by the slowest step which can be either mass transfer orreaction kinetics. For the purposes of copper electrowinning reactor design it is necessary todetermine the rate limiting step to optimize conditions so that capital costs and operatingability is optimized.

Heterogeneous electron transfer

By analogy with chemical kinetics for a simple first order reaction:

[13]

[14]

Using (5) the current for the forward reaction is given by:

[15]

And for the reverse reaction:

8/20/2019 213-240_Beukes

4/28


[16]

The total current density for the electrode is:

[17]

Using Arrhenius and the Activated Complex Theory it can be shown that rate of reaction kf and krev takes the forms:

[18a]

and

[18b]

E is the applied potential to the electrode and E° ′ the formal electrode potential that differsfrom the standard electrode potential by the activity coefficients. Recalling the NernstEquation:

[10]

The activity is equal to activity coefficient multiplied by the concentration in the bulk phase.Therefore:

[19]

It then follows that:

[20]

Equation [17] can be written as:

[21]

Substituting for kf and krev from equation [18] gives the Bulter-Volmer Equation (B-V):

[22]

This current-potential relationship governs all fast and single step heterogeneous electrontransfer reactions.

At equilibrium the exchange current density is:

[23]

The overvoltage (η) can be defined as:

[24]

8/20/2019 213-240_Beukes

5/28


where Eeq is the equilibrium voltage, Eeq=E o′when C ox(bulk) = C red(bulk) and E

o′=E

owhen

activity coefficients are equal to one, see Equation [20]. An expression for the equilibriumpotential is derived and shows Eeq to be close to the standard electrode potential and to varyaccording to changes in temperature and bulk concentrations.

[25]

The Butler-Volmer equation can then be written as follows:

[26]

This relationship shows that exponential changes to the current can result from changes to

the potential. Furthermore current is constrained by the surface to bulk concentration ratios of

oxidant and reductant species. The reaction rates do not grow indefinitely as potential is

increased and are thus limited by the transport of species to the electrode. A system that is

moved from equilibrium for Ox species to be reduced and Red species to be oxidized isdescribed by the B-V equation. This is achieved by setting the potential different to the

equilibrium potential, increasing the voltage thus increases the equilibrium difference which

increases the current hence speeding up the Faradaic process.

The maximum current that can be applied to maintain a reaction is known as the Diffusion

Limited Current. No matter what the standard rate constant is if the applied potential is

sufficiently large the maximum current will be reached. Assuming and adequate supply of

reactants to the reaction surface (the electrode) the rate of reaction is described by the Butler-

Volmer Equation. If the applied potential is adequate to maximize the Heterogeneous Electron

Transfer reactions the rate of reaction is then limited by the supply of reactants to the

electrode surface and is said to be mass transfer limiting (or controlled).

Assuming that the surface and bulk concentrations are equal (condition of non mass transfer

limited), for only large negative or positive overpotentials (only forward or reverse reaction

dominant) the B-V equation can be manipulated by taking a Log of both sides of the equation

then resolving for overvoltage gives:

[27]

And has the general form:[28]

Recognized as the well know Tafel Equation and is derived from the B-V equation for

specific condition of non mass transfer limiting, equal surface to bulk concentrations anddominant forward or reverse reactions. The procedure provides a means of linearizing the

relationship between overpotential and current or rate of reaction.

Mass transport

The movement of species from the bulk solution to the electrode surface occurs via three

possible mechanisms:

• Convection (or conveyance), forced or natural described by hydrodynamics ordensity/temperature differences

• Diffusion described by a gradient in concentrations

• Migration described by a gradient in electrical potential.

8/20/2019 213-240_Beukes

6/28


Experimental conditions of the electrode reactions are generally chosen to minimize the

effects of migration. This is done by providing a large quantity of inert electrolyte that does

not interfere with the electrode reaction, leaving only diffusion and convection mechanisms

for consideration.

Fick’s First Law for one-dimensional diffusion

[29]

Expanded to include convection (conveyance) and migration is the Nernst-Planck Equation8:

[30]

Minimizing the potential gradient using an inert electrolyte reduces the equation to:

[31]

This equation describes the one dimensional flux of species across the bulk solution to theelectrode interface due to the mechanism of conveyance and diffusion.

For a three dimensional volume it can be extended to:

[32]

Assuming that at steady state there is no change in concentration with time and that the

conveyance inside the diffusion layer is significantly smaller than the diffusion component,

Equation [32] becomes:

[33]

and by definition of rate of consumption

[34]

[35]

The rate of consumption (or generation) is equal to the rate change of concentration

difference across the diffusion layer. Using a circulation tank to minimize the change in

concentration over two of the three spatial axes we find that.

[36]

and

[37]

The use of the circulation tank to minimize the change in concentration over the two axes

parallel to the electrode surface significantly simplifies the mathematical mass transfer

relationship. Also sufficient concentration of ions across the face of the electrode is provided

to ensure that the mass transfer and supply of species to the surface of the electrode remains

sufficient. If the concentration gradients across the two parallel axes were not minimized a

8/20/2019 213-240_Beukes

7/28


varying current over the electrode surface would be required to maintain the rate of reaction.

The mathematical and practical difficulties of this are self evident. The resulting mass transfer

relationships are the following, and are relevant to one dimension normal to the electrode

surface.

[38]

And:

[39]

Faraday’s law:

[5]

Where:

[40]

Gives:

[41]

Substituting Equation [39] into equation [41] gives:

[42]

Defining the mass transfer coefficient (kd) as:

[43]

then

[44]

Equation [44] is different to Equation [42] in that using the mass transfer coefficient forces

an adequate hydrodynamic treatment of the flow reactor. Equation [42] is widely regarded in

the technical literature as the only mass-transfer equation for electrochemical systems. While

the difference may be trivial to the diffusion layer mechanism advocate not so from the

perspective of good chemical engineering practice, the universal application of heat and mass

transfer coefficients provides a more rigorous means of solving a hydrodynamic problem.

Mass transfer coefficients and diffusion limited current

Two equations that represent the same thing, namely, the reaction rate have been derived. The

B-V equation which describes Heterogeneous Electron Transfer and Ficks Law substituted

into Faradys Law that describes Mass Transfer. Increasing the overpotential up to a point

increases the reaction rate. When all species that reach the electrode are oxidized or reduced

the rate of mass transfer of the species to the electrode surface is the rate limiting step for the

production of copper. The diffusion limited current is the current above which an increase in

potential will not increase the rate of reaction. For that reason the Diffusion Limited Current

8/20/2019 213-240_Beukes

8/28


(DLC) should be determined to optimize the reactor design. In the field, operating reactors are

normally run at levels well below the DLC to achieve a good adherent product. This should be

considered when determining the actual applied current to the cell.

At the Diffusion Limited Current the surface concentration of species is zero meaning that

all surface species are consumed as quickly as they are supplied to the electrode. The DLC

equation then becomes:

[45]

Two important observations can be made at this point:

1) Increasing the concentration of the bulk reactant increases the DLC. This is a function

of the extraction process that was used to remove the copper from the host body and

other upgrade processes used before the electrowinning of the copper

2) Increasing the mass transfer coefficient increases the DLC. This is a function of the

hydrodynamics of the reactor cell and physical properties of the solution in which the

electrolyte is present such as temp, viscosity and other competing ions. The most

efficient supply of fresh solution is provided by enhancing the bulk motion. This iseasily achieved by stirring or by flowing the solution past the electrode or using other

methods to enhance the mass transfer coefficient.

Mass transfer coefficients are usually determined using empirical correlations that are based

on test work and made up of dimensionless parameters such as:

Reynolds Number:

[46]

Schmidt number:

[47]

Sherwood number:

[48]

Grashof number:

[49]

Prandtl number:

[50]

Rayleigh number:

[51]

A number of correlations appear in the literature and for convective mass transfer have the

general form:

[52]

8/20/2019 213-240_Beukes

9/28


For parallel plate electrodes of finite width and fully developed laminar flow. (Only

applicable if the maximum electrode length is less than 35 times the equivalent diameter. Re

8/20/2019 213-240_Beukes

10/28


kd3 m/s kd15,3 m/s kd4 m/s kd1 m/s

1 Nullabar M 2.09E-06 2.04E-06 1.76E-06 2.18E-06

2 Pasimco M 1.73E-06 1.68E-06 1.76E-06 2.11E-063 Girilambone Cu 2.08E-06 2.04E-06 1.76E-06 2.19E-06

4 CM Cerro Colorado 2.31E-06 2.28E-06 1.76E-06 2.18E-06

5 CM Quetrada Blanca 2.03E-06 1.98E-06 1.76E-06 2.27E-06

6 Minera El Abra 2.20E-06 2.16E-06 1.76E-06 2.20E-06

7 Codelco Chile Div 2.12E-06 2.09E-06 1.76E-06 2.18E-06

8 Empresa Minera 2.05E-06 2.02E-06 1.76E-06 2.18E-06

9 CM Zaldivar 2.18E-06 2.17E-06 1.76E-06 2.20E-06

10 Mantos Blancos 2.36E-06 2.36E-06 1.76E-06 2.11E-06

11 CM Carmen Andacolla 2.24E-06 2.21E-06 1.76E-06 2.21E-06

12 Hellenic Copper Mines 2.13E-06 2.11E-06 1.76E-06 2.18E-06

13 Nicico sarchesh 2.02E-06 1.98E-06 1.76E-06 2.12E-06

14 Mexicana de corbe 2.10E-06 2.07E-06 1.76E-06 2.05E-06

15 Miccl main plant 2.05E-06 2.01E-06 1.76E-06 2.17E-06

16 Southern Peru limited 2.22E-06 2.19E-06 1.76E-06 2.16E-06

17 Silver Bell Mining LLC 2.12E-06 2.08E-06 1.76E-06 2.22E-06

18 BHPB Copper San Manuel 1.95E-06 1.90E-06 1.76E-06 2.26E-06

19 Phelps Dodge Morenci S-side 1.79E-06 1.74E-06 1.75E-06 2.14E-06

20 Phelps Dodge Morenci Central 1.67E-06 1.61E-06 1.75E-06 2.16E-06

21 Phelps Dodge Morenci Stargo 1.85E-06 1.80E-06 1.75E-06 2.14E-06

22 First Quantum Bwana Mkubwa 1.94E-06 1.88E-06 1.75E-06 2.13E-06

i3 A/m2 i15,3 A/m2 i4 A/m2 i1 A/m2 i Faradaic A/m2

1 Nullabar M 240 235 203 250 226

2 Pasimco M 193 188 197 236 239

3 Girilambone Cu 246 241 209 260 321

4 CM Cerro Colorado 266 263 203 251 256

5 CM Quetrada Blanca 278 271 241 312 246

6 Minera El Abra 263 259 211 264 233

7 Codelco Chile Div 260 257 216 267 234

8 Empresa Minera 246 242 211 261 236

9 CM Zaldivar 285 283 230 287 244

10 Mantos Blancos 254 253 189 227 26211 CM Carmen Andacolla 275 271 215 271 286

12 Hellenic Copper Mines 260 256 214 265 272

13 Nicico Sarchesh 213 209 185 224 231

14 Mexicana de corbe 207 204 173 202 246

15 Miccl main plant 236 232 202 249 240

16 Southern Peru limited 253 250 201 247 266

17 Silver Bell Mining LLC 268 264 223 281 316

18 BHPB Copper San Manuel 258 251 233 299 231

19 Phelps Dodge Morenci S-side 205 199 201 245 234

20 Phelps Dodge Morenci Central 197 190 208 255 251

21 Phelps Dodge Morenci Stargo 210 205 200 243 239

22 First Quantum Bwana Mkubwa 205 199 186 226 196

Table I

Calculated kd (m/s) for 22 operating Cu EW plants

Table II

Calculated DLC (A/m2) for 22 operating Cu EW plants

8/20/2019 213-240_Beukes

11/28


The DLC should always be greater than the Faradaic current for a plant to ensure that a

good adherent copper cathode is produced and that the hydrodynamics of the system are

sufficient to ensure production is not limited by inadequate flow conditions.

It is important to note that the DLC’s reported here are based entirely on the correlations

used and not on any test work done by the authors. This method is used to demonstrate the

phenomenon of determining the DLC with common literature information, as well as thereasonably comparative results obtained from the different correlations. It can be seen that

some Faradaic currents are greater than the DLC’s calculated this does not mean that the plant

is incapable of adequate production but only means that the correlation may not be

representative of the exact conditions in that plant.

The correlations do not take into account gas evolving electrodes of which all copper

electrowining anodes are, this omission in the correlations may underestimate the DLC

significantly. This gas evolution can have a strong influence on the mass transfer of ions due

to the forced convection effects inside the parallel plate arrangement.

Overall electrode process

The earlier sections dealt with the two mechanisms of electron transfer for an electrode. This

section deals with an overall process of both heterogeneous electron transfer and mass transfer

that is relevant to copper electrowinning in practice.

The B-V equation that describes the Heterogeneous Electron transfer in terms of current and

overpotential is:

[26]

The current density is described by Fick’s and Faraday’s law for mass transfer is:[44]

And

[45]

Substituting and resolving for concentrations gives:

[60]

For copper reduction on an electrode the last term of the B-V equation can be neglected to

give:

[61]

Taking the log of both sides of the equation and separating for overpotential gives:

[62]

Substituting for the concentration ratio gives:

8/20/2019 213-240_Beukes

12/28


[63]

This equation describes overpotential, applied current and DLC in a single equation.

Example of fictional cell reactor

A fictional cell reactor is developed to show the relationship of overpotential to current forvarying conditions of concentration and flow in the cell. High level differences between thescenarios are:

• Cell 1, low flow rate• Cell 2, high flow rate• Cell 3, low flow rate high copper concentration• Cell 4, high flow rate high copper concentration.

Each cell has the same physical characteristics and features a cathode plating area of 1 m x 1m. The physical properties of the electrolyte are kept constant except for the concentrationsand flow rates in the cells. Table 3 also gives the DLC results for various correlations.

• The Forced Convection correlation of Equation [53] and Equations [54] described by thehydrodynamic characteristics of the cell are strongly influenced by the change in flowrate. Compared to the Natural convection correlations of Equations [55] and Equations[59] that, as expected, are not influenced by the changing flow rate as these correlationsdo not take into account the flow system but only density difference along the platesurface

• The graph of overpotential versus current is called a Voltammogram. For thedevelopment of the Voltammogram the correlation from Equations [53] is used as thisDLC changes with concentration and flow rate which is phenomenologically correct.Although it may not fully account for all mass transfer driving forces such as the gas

evolution at the anode this correlation demonstrates that reactor design is strongly

Cell 1 Cell 2 Cell 3 Cell 4

Fictional cell design units Lo MT Hi MT Lo MT Hi Cu Hi MT Hi Cu

Number of cathodes no off 48 48 48 48Number of anodes no off 49 49 49 49

Cell spacing A-C m 0.045 0.045 0.045 0.045

Cell inlet concentration g/l Cu2+ 40.00 35.77 38.75 36.07

Cell outlet concentration g/l Cu2+ 35 35 35 35

Cell volumetric flow rate m3/h 2.50 16.25 5.00 17.50

Copper production rate kg/h 12.5 12.5 18.75 18.75

Temperature °C 65 65 65 65

Density kg/m3 1216 1216 1216 1216

Diffusion coefficient m2/s 1.95E-09 1.95E-09 1.95E-09 1.95E-09

Current efficiency % 85 85 85 85

iL3 A/m2 191 300 228 309

iL15,3 A/m2 182 301 221 311

iL4 A/m2 273 244 264 246

iL1 A/m2 272 237 262 239

Faradaic current density A/m2 129 129 194 194

Table III

Operating conditions for the four scenarios compared

8/20/2019 213-240_Beukes

13/28


influenced by flow rate and concentrations.

The DLC3 for each cell is used to plot the voltammogram on overpotential versus currentdensity graph. The exchange current density io is taken from17 as

[64]

Figure 1 compares the Voltammograms for the four fictional cells:This graph shows the following conclusions to be drawn:

• Increasing the flow rate at the same copper production rate results in an increase of DLCfrom 190 A/m2 to 300 A/m2, ‘Cu Lo MT’ to ‘Cu Hi MT’

• Increasing the copper concentration and flow rate slightly results in increasing the DLCfrom 190 A/m2 to 230 A/m2, ‘Cu Lo MT’ to ‘Cu Lo MT Hi Conc’

• Increasing the copper concentration and flow rate significantly has the largest effect of increasing the DLC, from 190 A/m2 to 310 A/m2, ‘Cu Lo MT’ to ‘Cu Hi MT Hi Conc’.

Although the DLC equations used to determine this voltammogram are based on literaturework, the functionality remains correct. In order to optimize the cell design in terms of capitalcost for the same production rate, the DLC must be as high as possible so that the Faradaiccurrent density is as high as possible (limited to producing a good product). These conditionswill allow for the same transformer/rectifier arrangement to:

• Minimize installed electrowinning total plating area• Minimize installed electrowinning cell size• Minimize installed electrowinning cell house structure and civil footprint.

Methods to increase current density

Increased DLCs can be achieved by a number of different methods described in an excellentpaper on the subject10. Three main ways to increase the current density are suggested:

• Optimizing the cell design- EMEW Cell13

- Continental copper and steel (CCS) cell11

- Increased flow rate pattern and distribution in standard cell• Employing various types of forced convection

- Air sparging- Ultrasonic agitation

• Periodic current reversal.

Figure 1. Voltammograms for fictional cells

8/20/2019 213-240_Beukes

14/28


The EMEW cell is a recent design used to produce copper powder and circular plate at high

flow rates and high current densities. The CCS cell is a redundant design that used nozzles

placed at the cathode face to improve flow patterns.

The present paper does not discuss the various types of improvements that can be made but

does note that significant increases in current density can be obtained by using the methodslisted above. Table III compares operating conditions that produce a stable adherent pure

copper product at vastly increased current densities. The table is based on information from

reference10 as well as other sources.

The reported figures for the various techniques are significantly greater than standard

practice and design currently allows for. There are negative effects at these current densities

such as:

• Increased anode wear

• Acid mist control

• Corrosion of suspension bars

Cell potential

The total voltage across the cell can be divided into three components:

• The reversible decomposition potentials (Vmin)

• The activation and concentration overpotentials of the electrodes (ηa) (ηc).

• The Potential drop due to Ohmic resistance of the electrolyte and the electrical contacts

(V ohm).

[65]

The reversible decomposition potential is:

[66]

This is the difference between the two standard reduction potentials for the species being

oxidized and reduced and is the minimum voltage required at standard conditions

The overpotential due to potential and concentration has been discussed previously and are

represented by the modified B-V equation as:

[63]

And for no concentration overpotential influences then

[28]

Reported current densities A/m2

Type Low High

Traditional cell3,6,10,15,16 110 350

EMEW cell12,13 250 600

CSS cell11 430 1000

Air sparging10 1000 3000

Ultrasonic agitation10 3300 10500

Periodic current reversal10 430 ---

Table IV

Comparison of current density (A/m2) operating conditions of different cell types

8/20/2019 213-240_Beukes

15/28


The ohmic resistance due to the solution and the contacts are:

[67]

By Ohm’s law for a current flowing between the solution and an electrode. The relationship

between current drop and potential drop in an electrical conductor is5:

[68]

For current density in one direction only (parallel plates), Ohm’s law reduces to:

[69]

where the specific conductivity is a function of concentration and ionic mobility of the form5

[70]

The solution resistance and required potential difference is then a function of the

interelectrode gap, the ionic mobility, the concentration of the species in solution and the

current density applied to the solution.

Then for the electrowinning of copper from acid sulphate matrix:

[1]

[71]

[72]

Operated at a DLC of 300 A/m2 and an applied current density of 200 A/m2 (Cell 4

Conditions) the cathodic overpotential is:[73]

For the production of oxygen at the anode, excluding the acid concentration overpotential,

the Tafel equation is3

[74]

This gives:

[75]

The solution and contact resistance can be taken as

8/20/2019 213-240_Beukes

16/28


[77]

The efficiency is strongly dependant on additional side reactions that occur, for theelectrowinning of copper the decomposition of water lowers the current efficiency. Howeverthe single greatest effect is the presence of Iron in the solution. The cyclic oxidation andreduction of dissolved iron can significantly reduce the efficiency of an operation. Iron can beremoved by a number of methods that include.

• Solution precipitation• Prior reduction of iron (III) by SO2 or copper metal• Increasing the bleed stream volume• Use of a diaphragm cell.

Current efficiencies for direct electrowinning operations can be as low as 65% whileefficiencies for post SX electrowinning can be as high as 95%.

Part 2: Practical considerations

The following is based on a case study of the Ruashi BMR designed by TWP Matomo

Process Plant

Production rates

Production rates are usually specified by the client, depending on the mass balance and theenvironment in which the electrowinning is to be designed for. Design conditions willnormally be in the region of specified rate plus 10%, this excludes availability.

Availability

Mechanical availability depends on the following factors:

• Site conditions and location. The Ruashi BMR is located in Lubumbashi, DRC, and assuch, consideration was given to have stand-by units for all key mechanical equipment.Operation within South Africa would not necessarily require the same design. Dualavailability design was included for cellhouse circulation pumps, feed to and return frompumps servicing the cellhouse, overhead cranes, stripping machines (for which anadditional allowance was made to allow manual stripping of cathodes in a specificallydesignated area in case of equipment breakdown). Both the loaded electrolyte pondfeeding the cellhouse as well as the spent electrolyte pond returning spent electrolytesolution to the solvent extraction circuit is of split design, providing for maintenance on

the pond linings without having to shut down the plant• Maintenance programmes proposed or in place at the operation depend on sitemanagement in terms of pro-active maintenance, service schedules, etc. This cannotalways be quantified, especially for new operations; however, in general, largerestablished companies normally have a better maintenance system in place than smallcompanies

• Education and experience level of both maintenance and operational personnel. Considergeographic location of the area, whether there are existing operations from wherepersonnel could be sourced, language barriers, culture of work, political stability.

Current density

Depending on the type of operation, current densities can vary between 200 and up to about

8/20/2019 213-240_Beukes

17/28


375 A/m2. Operations where Solvent Extraction precedes Electrowinning normally produce

cathodes with a good surface quality at higher current densities than direct electrowinning

operations. Furthermore, the clarity of the feed solution needs to be taken into account. The

Ruashi design includes multimedia coalescing filters for the removal of both solids as well as

organics from the Cu Feed solution before the solution reports to the Cellhouse. Process

Engineers are in general quite comfortable operating in the region of 250 to 300 A/m2 whenSX and multimedia filters are employed.

It is also advisable to consult the vendors of the rectiformers. There are standard designs for

the transformer-rectifier pairs, ranging in size from small up to 30 kA. For the Ruashi design,

two sets of 30 kA were purchased to satisfy the design requirement of 60 kA total DC supply

to the electrowinning cells. Depending on the rectiformer supply, the cell design can be

modified to suit.

Cathode centre to centre spacing

Design current density is also dependant on electrode spacing. The minimum spacing

recommended for Cu Electrowinning operations for cathode centre to cathode centre is 95mm. Below this the risk of shorts due to electrode alignment and nodular growths is

considered to be unacceptably high.

Cathode quality

This is normally specified by the client. Within the Cellhouse, there are very few variablesthat influence cathode quality; it is normally a function of the preceding purification steps.

Sampling specification and procedure

Sampling on site is normally done on site for in-house quality control purposes only. The

client usually sells his product through a third party, who will be responsible for sample

analysis. Some mining houses have accredited laboratories, in this case, the laboratoryanalyze samples received from the plant. Two commonly used methods to obtain a sample are

to either drill holes in randomly selected samples on randomly selected locations of the

cathode. The drillings are then sent for analysis in the accredited laboratory. The second

method follows the same procedure, however instead of drillings; a manually operated

punching machine is used. It has to be taken into consideration that normal drill bits are not

used for the drilling of samples, since iron contamination of the samples frequently occurs.

Regarding the selection of cathodes in a group to be analyzed and the location on each

cathode, a standard procedure was published by the ISO organization. The standard title is

‘ISO 7156 – Refined Nickel – Sampling’. No procedure for Copper sampling could be found

for Copper Sampling, however the same principles for Cu sampling as for Ni sampling is

assumed to be applicable.

Number of cells

This is dependant on a number of factors, including cell dimensions, real estate available forthe Cellhouse building, applied current density, production rate, DLC, current efficiency.There is normally an equal amount of cells, making the busbar arrangement practical. Cells ina Leach-Solvent Extraction-Electrowinning design should be split into scavenger andproduction cells to limit organic contamination of cathodes in case of organic carry-oversoccurring at the Solvent Extraction Purification Plant. Scavenger Cells are then designed tomaintain the same cathode face velocity as for the Production/Commercial Cells of

approximately 0.08m3

/hr/m2

cathode surface, the difference being that there is a once-off pass

8/20/2019 213-240_Beukes

18/28


of electrolyte through these cells. The scavenger cells then produce the potentially poorsurface quality cathode by burning off any organic substances, before the bulk production of cathodes in the commercial cells. For Ruashi, the design accommodated 24 scavenger cells,which constituted exactly one bank of cells. This design facilitated easy piping arrangementof feed and overflow to the cells.

Cellhouse layout

Practical considerations to be considered include optimizing availability and maintaining

flexibility inside the Cellhouse. For Ruashi, the Cellhouse was split into two sides, each

containing two banks of 24 cells. This configuration allowed the two stripping machines to be

placed in the middle of the Cellhouse, minimizing travel time for the cranes harvesting the

cathodes and returning blank cathodes to the cells. Furthermore, it allows one half of the cells

to be shorted out with relatively little effort, which came in handy during the commissioning

of the Cellhouse. The connecting busbar from the end of the Cellhouse was connected to the

centre of the Cellhouse, allowing a ramp-up stage of electrowinning on the one side, while

construction was still ongoing on the other side of the Cellhouse.

The distance between the end of the cell and the supporting structure of the Cellhouse has tobe taken into consideration in order to minimize the possibility of electrical accidents. A

distance of at least 2 to 3 metres is recommended. The same applies to the distance between

the cells comprising one bank and the opposite bank of cells (the ‘middle’ of the Cellhouse

walkway).

Feed valves to the individual cells should be practically located in order to allow easy

operation of these valves, without interfering too much with the harvesting procedure. For

Ruashi, the feed valves were located on top of the cells in the middle of the Cellhouse

walkway.

Walkways need to be constructed from non-conductive material. Two options are available,

the one being wooden walkways, of which a high grade knot-free pine is considered the most

feasible option. Wood is then also treated with CCA (Copper/Chrome/Arsenic). The secondoption is the use of FRP grating. This is much more expensive, but operating cost is much

lower. Certainly, for the areas close to the liquid, FRP will have to take preference, since the

timber will not last in this environment. For Ruashi, the majority of the walkways consist of

CCA treated timber, with some FRP inserts at the two ends of the cell to allow for visual

inspection of overflows.

Figure 2. Cell house layouts showing access platform, feed pipes

8/20/2019 213-240_Beukes

19/28


Figure 4. Showing internal feed pipe arrangement

Figure 3. Showing Intercell busbar and lead anode

Figure 5. Showing Cathode bail removing four cathodes

8/20/2019 213-240_Beukes

20/28


For practicality, the Cell Feed inlet pipes all run along the centre of the Cellhouse, allowing

for only two feed manifolds to feed the entire Cellhouse, that is, both the East and the West

Banks. The feed manifolds enter the Cellhouse at the electrical neutral point in case of the

production cells, and enters in the middle of the Cellhouse for the scavenger cells. This was to

accommodate piping where limited space was available between the two banks of cells.

Ideally, the scavenger cells’ feed manifold would also have entered the Cellhouse at theelectrical neutral point to minimize stray currents. Cell drain points were also located in the

centre of the building, thus allowing the civil slope of the floor to act as a run-off trench for

spillages. Spillages are thus mainly confined to only the centre of the Cellhouse, minimizing

acidic attack of the civil structures to only the centre of the building.

Piping layout

For practicality, the cell feed inlet pipes all run along the centre of the Cellhouse, allowing for

only two feed manifolds to feed the entire Cellhouse, that is, both the East and the West

Banks. The feed manifolds enter the Cellhouse at the electrical neutral point in case of the

production cells, and enters in the middle of the Cellhouse for the scavenger cells. This was to

accommodate piping where limited space was available between the two banks of cells.Ideally, the scavenger cells’ feed manifold would also have entered the Cellhouse at the

electrical neutral point to minimize stray currents. Cell drain points were also located in the

centre of the building, thus allowing the civil slope of the floor to act as a run-off trench for

spillages. Spillages are thus mainly confined to only the centre of the Cellhouse, minimizing

acidic attack of the civil structures to only the centre of the building.

Cell elevation

In order to facilitate the gravity overflow of cells to the Circulation Tank, cells need to be

elevated of the ground. When elevation is already a necessity, it is then useful to elevate the

cells high enough to accommodate access for personnel to inspect the civil structures fromtime to time, especially since the Cellhouse is such a corrosive environment. Additionally,

floor space underneath the cells can then also be used as storage space for cell furniture. The

Ruashi Cells were lifted approximately 1,6 m of the ground. Even so, the terrace level of the

Tankfarm had to be lowered to allow overflow into the Circulation Tank. Additionally, the

Circulation Tank was designed low with a high floor space (3,5 m high by 12,5 m diameter) to

further accommodate gravity feed into the tank.

Cell potential and current efficiency

Cell potential is a function of the respective half reaction potentials, losses through

electrolyte, busbars and competing reactions. It is normally in the range of 1,9V to 2,3V,

lower for the purer Leach-SX-EW operations than for the direct electrowinning operations.

Current efficiency for direct electrowinning operations can be as low as 65%, and for Leach-

SX-EW operations up to about 93%. It depends on the concentration of Fe in the feed

solution, as well as proper housekeeping to ensure good electrical contact between electrode

hanger bars and the triangular busbars.

Cathode face velocity

A number ranging between 0,05 and 0.1m3/hr/m2 of cathode surface area available for plating.

It influences the surface quality of the cathode by breaking down the diffusion layer, but too

high a face velocity can cause ‘flushing’ of the cathode surface area, creating an area of no

plating where the feed is introduced into the cells. For Cu Electrowinning operations, a face

8/20/2019 213-240_Beukes

21/28


Figure 8. Showing cathode stripping machine

Figure 7. Showing Copper busbars

Figure 6. Overflow launders between cells

8/20/2019 213-240_Beukes

22/28


velocity of 0.08 is normally employed. The cathode face velocity will then determine the flow

rate per cell, and thus the internal recycle required inside the Cellhouse.

Feed and spent tenors

The Leach and Solvent Extraction processes will determine the copper and acid tenors of the

feed to the Cellhouse. For Leach-SX-EW operations, feed tenors of Cu can go as low as

25 g/l Cu, depending on the applied current density, which will have to be reduced as the Cu

tenor is reduced. The sulphuric acid tenor of the feed solution is a function of the stripping

process inside SX, and that of the spent solution is a function of the rate of copper production,

where a stoichiometric balance exists of 1,98 g of acid produced for every gram of copper

plated.

The Ruashi EW operation was started up with unpurified Primary Leach Solution at a pH of

~1,8 and a Copper tenor of 24 g/l , the resultant cathode had a very good surface quality, albeit

at an applied current density of only 75 A/m2. Raising the current density to 110A/m2 resulted

in the formation of nodules, but this was attributed to no control over the flow rate (cathode

face velocity) to the cells. Since the flow rate was increased, good surface quality cathodes areachieved at current densities of 100A/m2 and Cu tenors of ~27 g/l . Ruashi design caters for a

Cu Feed tenor of 50 gpL and a Cellhouse bite of 5 g/l Cu, at these operating conditions; the

design applied current density of 275A/m2 is expected to deliver good surface quality

cathodes. This has not yet been achieved at Ruashi, due to the SX section still being

constructed and the front end of the plant still being commissioned.

Cathode harvesting

The cathode harvesting cycle was designed so as to accommodate a deposit thickness of

~5mm. The thickness is dependant on the capability of the stripping machines to separate the

copper from the blanks, for which the stripping machine suppliers have to be consulted.Furthermore, the longer a cathode is allowed to plate in a cell, the higher the risk of shorting

becomes due to the exponential nature of nodular growths inside the cell. A harvesting cycle

of more than eight days is not recommended, but this is again dependant on the applied

current density (rate of plating).

Busbar design

In order to minimize heat losses through the busbars, the current density through the busbars

have to be limited. Conservative clients prefer not to go above 1A/mm 2, while design

companies are normally quite comfortable designing busbars at 1,2A/mm2. Of importance is

to run parallel busbars with a gap between them so as to allow heat reduction through aircooling. The Ruashi busbar designs for the main busbars were 6 busbars, each of size 450 mm

x 20 mm, separated by a 20 mm gap between each busbar. This fulfilled the 1,2A/mm2

requirement for the total design current of 61kA.

Intercell busbar design

Two options are available for the design of intercell busbars. Conventional dogbone type of

designs was previously employed and is still preferred by some clients. It is however much

more expensive than triangular busbars, due to the larger amount of copper needed. Triangular

busbars are a more viable option, provided that proper consideration is given to the design and

layout of contact points between hanger bar (cathode and anode) and triangular busbar. It

8/20/2019 213-240_Beukes

23/28


becomes even more critical when the design had to cater for cathodes that are symmetrical to

cater for the 180o rotation during the stripping operation. Furthermore, due consideration

needs to be given to the different elevations between cathode and anode hanger bars on the

cell to allow for shorting out of the cell with a shorting frame. Although this can be avoided if

a bullhorn anode hanger bar design is employed, it is unnecessary and a straight anode hanger

bar design allows for the maintenance of anodes using the same bailer as is used for cathodeharvesting.

Cathode design to cater for hooking by the bailer can be through either cut-outs from the

cathode surface above the liquid level, or through the attachment of stainless steel fish-eyes to

the hanger bars. Fish eyes are economically more feasible since it comprises less work for the

suppliers. It then implies that cut-outs are made from the anode surfaces to accommodate the

use of one bailer size to both harvest cathodes as well as do anode maintenance.

Dimension and number of electrodes per cell

Cathode size is selected to accommodate industry standards, where shipping of cathodes in

containers as well as furnace openings in the production industry has to be taken into

consideration. Client liaison will normally give a good guideline as to cathode sizing. Anodes

are designed at least 30 mm bigger than cathodes on both the width and the height (wetted

dimensions). This is to ensure that copper plating does not occur around the sides of the

cathode. There will always be one more anode than cathode per cell, and depending on the

harvesting design, either every second or every third cathode is harvested per pull. Using the

standard design of every third cathode, cathodes per cell then have to equal multiples of three.

Anode composition and type

New electrowinning operations at the moment almost exclusively employ cold-rolled lead

anodes, since it yields better dimensional stability. Standard additions to the lead anodeincludes a fraction of Calcium (0,05 to 0,08%) and Tin (1,2 to 1,5%). Anode life expectancy

is in the order of 7 to 9 years.

Electrode furniture

Anode buttons from non-conductive PVC material is readily available on the market. Three

to five anode buttons are located to the anode, that being in the two bottom corners of the

anode, one in the centre of the anode, and an optional two at the two top corners of the anode.

The anode buttons prevent shorts between cathode and anode. Edge strips are located to the

edges of the stainless steel cathodes to ensure that the two sides of the cathode are not

intergrown. Various types of edge strips are available. Ruashi and various EW operations inZambia employ the Rehau ‘cross-slot’ configuration of edge strips. It is important to design

the edge strips long enough so that it ends well above the liquid level of the cell, since liquid

between the edge strip and the cathode will cause metal plating and thus damage to the edge

strips.

Electrowinning cells

Continuous improvement to the polymer concrete industry has led to the almost exclusive use

of polymer concrete as material of construction for electrowinning cells. There are two main

suppliers in Southern Africa, they being CSI and PCI. Product quality from both suppliers is

acceptable and the Copper industry in Zambia is split almost equally between the two

8/20/2019 213-240_Beukes

24/28


suppliers. The polymer concrete cells contain a fiberglass lining, which is part of the

moulding/casting process. This further protects the concrete from acid attack. Other

operations use lead linings or normal concrete with fiberglass lining moulds to house the

electrowinning process. This is not feasible for new operations.

Cathode strippingVarious suppliers can custom design stripping machines for the electrowinning operation.

Options include fully automatic machines, semi automatic machines and options for water

spray rinsing booths. Design considerations should include simplicity if the machine is used

in remote locations, availability of spares, stripping rate and capital expenditure. Ruashi

employs two semi-automatic Styria stripping machines; this gives flexibility in case of a

breakdown as well as ease of maintenance.

Plating agents

Various smoothing agents are commercially available, ranging from glues to guars to

synthetic products. It is best to test these in research facilities such as at Mintek to determinethe effect of the smoothing agent on current efficiency and surface quality of the cathode.

Dosage rates range from 150 to 400g of smoothing agent per ton of Cu produced. Chloride

levels inside the Cellhouse should be kept below 30ppm to ensure the stainless steel cathodes

are not corroded. Addition points can be at the Circulation Tank for the Commercial Cells and

inline at the Feed manifold for the Scavenger Cells. Care should be taken to design the

Scavenger system so as not to have reverse flow of electrolyte back into the smoothing agent

supply tank, as such a non-return valve is recommended in-line for the scavenger cells

smoothing agent supply.

Acid mist management

Hollow polypropylene balls are commonly employed in Cellhouses. Additionally, synthetic

foaming agents are available, the effect of these on the Solvent Extraction should be

considered. Designing a Cellhouse with open sides helps with natural ventilation, and if

CAPEX considerations are not of primary importance, extraction fans can be employed to

minimize acid mist.

Temperature management

Feed to the Cellhouse could be heated using heat exchangers, the optimum temperature is

considered to be in the range of 45 to 55°C for optimum electrolyte conductivity. Due to heat

generation inside the Cellhouse, a cooling heat exchanger is then required for both circulating

electrolyte and spent electrolyte. Solvent Extraction plants which use Spent electrolyte forstripping need to be maintained below 40°C to prevent the degeneration of the organic phase.

Inventories

For optimum plant availability, sufficient storage capacity is required. Ruashi employs split

ponds with double liners and leak detection systems. In designing the storage pond systems,

consideration need to be given to the type of pumps to be used. Options available include self-

priming pumps (Sulzer) or conventional pumps with priming tanks. The self-priming pumps

employed at Ruashi have proved to be difficult to commission, and to date little support from

the supplier was received. Additionally, self-priming pumps are considerably more expensive

than the combination conventional pump and priming tank system.

8/20/2019 213-240_Beukes

25/28


Cathode washing

Heated dip tanks are allowed for at the Ruashi BMR, however other systems available include

the use of heated water spray systems or, as previously mentioned, spray systems as part of a

stripping machine.

Electrolyte filtration

Scheibler filters are employed in industry to aid in the filtration of circulating electrolyte,

there-by reducing the need to clean cells. Capital expenditure availability has to be

considered, and if a filter is not part of the initial design, real estate allowance should be made

to later locate the filter when client finances allows.

Conclusions

The above paper has two parts, the first part describes the theoretical requirements for copper

electrowinning and the second part describes the practical requirements based on the Ruashi

case study.The first part is based on the fundamental equations of reaction rate kinetics and

thermodynamics as well as hydrodynamics to describe the baseline equations that can be used

in designing a copper electrowinning process. The important considerations for the designer

are determination of the Diffusion Limited Current and the appropriate selection of the

operating current. These numbers which are based on numerous variables such as flow rate to

cell, cell bite, plating area, copper, acid and iron concentrations, temperature of electrolyte

and current efficiency will determine the structural size and footprint of the copper

electrowinning process building. They have the largest effect on overall capital cost and

consequently the timeline of building the process plant. It has been discussed that the use of

high current densities requires some modifications to the conventional 100 year oldelectrowinning cell design. Conventional cells are operated at maximums of approximately

~350-400A/m2 but most designers will not exceed these values in conventional designs, with

good reason. With adequate hydrodynamics and increase of the Diffusion Limited Current the

capital cost of a copper electrowinning building can be reduced. There are problems

associated with this, particularly acid mist generation. Adequate ventilation systems or

enclosed hoods similar to those used in the nickel industry can be used. Any designer straying

from the path of tried and tested technology should confirm the new design with appropriate

test work before committing the clients’ project and their career to the annals of technical

success or failure.

The second part of the paper discusses practical requirements around the electrowinning

building and process areas based on the Ruashi case study. A significant number of issues

need to be addressed during the design phase to ensure a successful project and an operable

plant. Key aspects such as plant location and skill set are critical. Physical aspects such as

electrolyte storage, operating temperatures, cell sizes, busbar design, circulation tank and

piping layout are very important. Plating and leveling agents need to be considered. Cathode

product handling and acid mist control are vital.

Acknowledgement

The authors would like to thank Metorex the owners of Ruashi BMR for the opportunity to

present their findings.

8/20/2019 213-240_Beukes

26/28


Nonmenclature

Symbol Definition SI Units

Q Charge Coulombs

I Current Amps, C/s

n Valance gEq/gmolF Faraday’s constant C/gEq

M Number of Mols gmol

i Current density A/m2

A Area m2

dQ/dt Charge per time C/s

dM/dt Mols per time gmol/s

G Gibbs free energy J/gmol

Go

Standard Gibbs free energy J/gmol

R Universal gas constant J/(gmol.K)

T Temperature KelvinaP

nActivity of products Dimensionless

aRm

Activity of reactants Dimensionless

aRed Activity of oxidized species Dimensionless

aOx Activity of reduced species Dimensionless

V Potential Volts

W Work Joule

E Potential Volts

kd Mass transfer coefficient m/s

kr Reaction rate constant 1/s

Ox Oxidized species Dimensionless

Red Reduced species Dimensionless

e-

Electrons Dimensionless

kf Rate constant of forward reaction 1/s

Cox Concentration of oxidized species gmol/m3

ic Current density of cathodic reaction A/m2

ia Current density of anodic reaction A/m2

krev Rate constant of reverse reaction 1/s

Cred Concentration of reduced species gmol/m3

ko Rate constant at equilibrium 1/s

α Separation factor Dimensionless

Eo’

Formal electrode potential Volts

Eo

Standard electrode potential Volts

γ Activity coefficient Dimensionless

io Exchange current density A/m2

Eeq Equilibrium potential Volts

η Overpotential (overvoltage) Volts

Nox” Mass flux of oxidized species gmol/(m2.s)

D Diffusion coefficient of species m2/s

(x,t) Species vector at time t and position x -------

8/20/2019 213-240_Beukes

27/28


v Velocity of species m/s

dΦ/dx Potential gradient across electrolyte Volts/m

∇ Laplace operator -------

ra•

Reaction rate for species a gmol/(m3.s)

rox•

Reaction rate for oxidized species gmol/(m3.s)

rred • Reaction rate for reduced species gmol/(m3.s)

dx,dy,dz Distance in x,y and z directions m

δ Boundary layer thickness m

iL Diffusion limited current A/m2

Re Reynolds number Dimensionless

L Characteristic length m

ρ Density kg/m3

µ Viscosity kg/(m.s)

Sc Schmidt number Dimensionless

Sh Sherwood number Dimensionlessde Characteristic length/hydraulic diameter m

Gr Grashof number Dimensionless

g Gravitational acceleration m/s2

β Thermal expansion coefficient 1/K

Ts Temperature at the surface K

T∞ Temperature at bulk K

Pr Prandtl number Dimensionless

Cp Specific heat capacity J/(kg.K)

k Thermal conductivity W/(m.K)

Ra Rayleigh number Dimensionless

Nu Nusselt number Dimensionless

Le Lewis number Dimensionless

Va Anodic standard electrode potential Volts

Vc Cathodic standard electrode potential Volts

Vmin Decomposition potential Volts

Vohm Ohmic resistance Volts

Vir Solution potential Volts

Vcontacts Contacts potential Volts

κ Resistivity S/cmui Ionic mobility of species i gmol.cm2/(J.s)

ε Current efficiency percentage

References

1. HAYES, P.C. Process Principles in Mineral and Materials Production, Hayes PublishingCo., 1983.

2. PETRUCCI, R.H. AND HARWOOD, W.S. General Chemistry Principles and Modern Applications, Prentice Hall International, Inc., 1993.

3. LORENZEN, L. Mineral Processing Class Notes, Electrochemical Reactor Design,

University of Stellenbosch, 2000.

8/20/2019 213-240_Beukes

28/28


4. INCOPERA, F.P. and DEWITT, D.P. Fundamentals of Heat and Mass Transfer, JohnWiley and Sons, Inc., 1996, (4th Edition), Chap 6, Chap 9.

5. PERRY, R.H. and GREEN, D.W. Perry’s Chemical Engineers’ Handbook , McGraw-Hill,1997.

6. JACKSON, E. Hydrometallurgical Extraction and Reclamation, Ellis Horwood Limited,

1986, Chap 5.7. WANG, J. Analytical Electrochemistry, John Wiley and Sons, Inc., 2001, (2nd Edition).

8. ZANELLO, P. Inorganic Electrochemistry: Theory, Practice and Application, The RoyalSociety of Chemistry, 2003.

9. BRETT, C.M.A. and BRETT, A.M.O. Electrochemistry Principles, Methods, and Applications, Oxford University Press, Inc., 1993.

10. MACKINNON, D.J. and LAKSHMANAN, V.I. Recent Advances in CopperElectrowinning, Mineral Research Program, Mineral Research Laboratories, CANMETReport 76–10, 1976.

11. ANDERSEN, A.K. and BALBERYSZSKI, T. Electrowinning of Copper at High Current

Densities with the CCS Cell, The Metallurgical Society of AIME, TMS Paper Selection,Paper A68-17, 1968.

12. ESCOBAR, V., TREASURE, T., and DIXON, R.E. High Current Density EMEW®

Copper Electrowinning, Electrometals Technologies Ltd. Official Website.http://www.electrometals.co.au.

13. ROUX, E., GNOINSKI, J., EWART, I., and DREISINGER, D., Cu-Removal From theSkorpion Circuit Using EMEW® Technology, The South African Institute of Mining andMetallurgy, The Fourth Southern African Conference on Base Metals, 2007.

14. Short course on Electrochemistry, Electrochemical Engineering and Electrometallurgy:Module on Thermodaynamics and Kinetics, University of the Witwatersrand,Johannesburg, 2005.

15. SANDENBERGH, R. University of Pretoria Electrochemistry Class Notes, 2007.16. HOULACHI, G.E., EDWARDS, J.D., ROBINSON, T.G. Cu 2007, vol. V, Copper

Electrowinning and Elecrorefining, Toronto, Metsoc Publication, August 2007.

17. Short course on Electrochemistry, Electrochemical Engineering and Electrometallurgy:Module on Applications of Fundamentals to Electrowinning and Electrorefining of Metals. University of the Witwatersrand, Johannesburg, 2005.

Nicholas Terence BeukesProcess Engineer, TWP Matoma Process Plant, South Africa

Nick has worked as a process engineer for seven years since

graduating as a chemical engineer. Experience includes projects

commissioning for Anglo Platinum copper-nickel matte smelting and

converting. Following that he has worked as a process design engineer

involved primarily in mass and energy balance, pipe and

instrumentation, process flow diagrams and equipment selection and

sizing, mechanical layout and functional specification design and development. The above

includes ore beneficiation, hydrometallurgy and pyrometallurgy applications, such as chrome

recovery, gold, uranium and base metal refining.

Documents

213-240_Beukes